# Instrument Identifier AI Detect which instruments play in a track: onset-split note analysis scores every event against 13 timbre profiles and reports ranked instruments, families and per-note evidence. > Canonical page: https://elysiatools.com/en/tools/instrument-identifier-ai - **Category:** Media - **Keywords:** instrument recognition, instrument identifier, timbre analysis, music instrument detection, which instrument, audio analysis, harmonic analysis, mpeg-7, log attack time, inharmonicity, vibrato detection, instrument family ## Overview This is a signal-analysis identifier, not a magic black box. Every detected note event is measured the way timbre research measures it: MPEG-7-style log-attack-time (time from 2% to 100% of peak energy, Peeters 2000), harmonic power ratio, the odd/even partial power ratio over harmonics 1-5 that separates a clarinet’s cylindrical bore from a trumpet’s flare (Almeida 2024), the string-stretch inharmonicity coefficient B from fₙ = n·f₀·√(1+B·n²) that fingerprints piano vs guitar strings (Fletcher 2000), spectral centroid/spread/flatness, and a 4-7 Hz vibrato detector that is the classic vocal/violin signature. Each event is scored against 13 profiles (piano, guitar, bass, violin, flute, clarinet, sax, brass, organ, synth lead/pad, voice, drums) and rolled up into instrument families. Be realistic about polyphonic mixes: state-of-the-art classifiers reach roughly 35% per-instrument and 77% per-family accuracy on isolated notes (Eronen 2001), so treat results as a strong hint, strongest on exposed solos, stems and one-instrument loops. ## Inputs - **Music track** (file) - **Analysis mode** (select) - **Minimum note length (ms)** (number) ## When to use - Analyzing isolated instrument stems, solo passages, or single-instrument sample loops to identify the source instrument. - Auditing woodwind, brass, bowed string, and percussive timbre characteristics in unlabelled audio recordings. - Inspecting per-note acoustic properties like harmonic distribution, log-attack time, and vibrato frequency across an audio track. ## How it works - Upload an audio file up to 100 MB and select whether to run a segmented per-note onset analysis or a whole-track analysis. - Adjust the minimum note length threshold (between 80 ms and 500 ms) to configure how short transients and sustained notes are detected. - The engine measures spectral centroids, attack times, odd/even harmonic ratios, string inharmonicity coefficients, and vibrato signatures on each note event. - Review an interactive HTML report displaying ranked instrument predictions, instrument family breakdowns, and a per-note acoustic timeline. ## Use cases - Sample library cataloging: Automatically tagging unlabeled instrument loops and one-shot acoustic recordings by family and timbre profile. - Music education and ear training: Visualizing harmonic partial structures, attack times, and inharmonicity differences between similar instruments like clarinet and trumpet. - Stem verification: Checking exported audio stems in a production session to confirm solo instrument tracking and pitch-timbre consistency. ## Frequently asked questions ### Which 13 instrument profiles can this tool recognize? It evaluates audio against profiles for piano, guitar, bass, violin, flute, clarinet, saxophone, brass, organ, synth lead/pad, voice, and drums. ### How accurate is the tool on complex, polyphonic full mixes? Accuracy is highest on isolated notes, solo stems, and single-instrument loops. Full polyphonic mixes produce overlapping harmonics that lower classification precision. ### What is the difference between per-note and whole-track analysis? Per-note mode splits audio by detected onsets to measure each note's attack and timbre independently, while whole-track mode averages spectral metrics across the entire duration. ### What does the minimum note length setting do? It sets the shortest duration (in milliseconds) a detected sound event must sustain to be classified, filtering out accidental clicks and brief transients. ### What audio formats and file sizes are supported? The tool accepts standard audio file formats with a maximum file upload limit of 100 MB. ## Related tools - [Genre Classifier AI](https://elysiatools.com/en/tools/genre-classifier-ai): Classify a track into genres from measured evidence: GTZAN timbral-texture features, beat-grid pulse shape, band balance, key and dynamics — with a ranked top-3 and the reason behind every pick. - [Mastering Chain Pro](https://elysiatools.com/en/tools/mastering-chain-pro): Run a full music mastering chain on one track: sub-rumble high-pass, corrective EQ, glue compression, character EQ, saturation, stereo widening, true-peak limiter, and two-pass linear loudness normalization to 2026 platform targets. - [Podcast Chapter Marker Builder](https://elysiatools.com/en/tools/podcast-chapter-marker-builder): Build every podcast chapter format from one timecoded list: Podcasting 2.0 JSON + RSS tag, ID3v2.4 CHAP+CTOC burned into an MP3, Vorbis comments, mp4chaps, YouTube timestamps and SRT, with a per-player support matrix. - [Karaoke Vocal Reducer](https://elysiatools.com/en/tools/audio-karaoke-track): Attempt to remove vocals from a stereo track - [Audio Loudness LUFS Normalizer](https://elysiatools.com/en/tools/audio-loudness-lufs-normalizer): Measure integrated LUFS, true peak, and LRA, then normalize audio to Spotify, Apple Music, broadcast, or custom targets - [Audio Click Removal](https://elysiatools.com/en/tools/audio-click-removal): Remove clicks and pops from vinyl recordings - [Rhythm Extractor](https://elysiatools.com/en/tools/rhythm-extractor): Analyze a drum track or full mix into a step-sequencer pattern: BPM, beat grid, kick/snare/hi-hat hits quantized per bar, drum MIDI, and an HTML grid report. - [Convert GIF to Raw Pixel Buffer](https://elysiatools.com/en/tools/gif-to-raw): Export GIF frames as raw pixel buffer data for analysis, rendering pipelines, and low-level image processing. ## Samples - [Copyright-Free FLAC Audio Samples](https://elysiatools.com/en/samples/flac-samples): Lossless FLAC audio samples for testing and development, mirrored from MP3 set with nature sounds and meditation music - [Copyright-Free MP3 Audio Samples](https://elysiatools.com/en/samples/mp3-samples): Collection of royalty-free audio samples for testing and development purposes including nature sounds, meditation music, and ambient audio - [Copyright-Free WAV Audio Samples](https://elysiatools.com/en/samples/wav-samples): Uncompressed PCM WAV audio samples for testing and development, mirrored from MP3 set with nature sounds and meditation music - [Copyright-Free Raw PCM Audio Samples](https://elysiatools.com/en/samples/pcm-samples): Raw PCM s16le audio samples featuring nature ambience and relaxing music for waveform, playback, and conversion workflows